Agenda

• brief introduction to ::
• ggplot() flyover
• remarks about facets
• color vs. fill

Reminders

• Midterm is October 21st in class
  • Review is posted on Canvas
  • Let me know if you have any questions or want solutions to previous assignments
  • I will try to have all grading done by October 17th
• Final Project Ideas due October 28th
  • need a “solid” idea of what you might look at
  • need data sets, or ideas for data sets
    • these do not need to be your final data sets, but I just want to see you have been thinking about this project
• Groups Due November 2nd
  • You can work by yourself or with a partner
  • Fill out the canvas “quiz” to let me know if you want to work by yourself or with someone else.
  • If you want to work with a partner, but don’t know who to work with, let me know and I will find someone

PackageName::FunctionName

• We can use the :: function to reference functions inside packages.
  • This helps us be extra sure that we are using the exact function we want to be
  • Helps avoid conflict (i.e. when 2 packages that have a function with the same name that do different things)

library(tidyverse)
head(diamonds)

## # A tibble: 6 × 10
##   carat cut       color clarity depth table price     x     y     z
##   <dbl> <ord>     <ord> <ord>   <dbl> <dbl> <int> <dbl> <dbl> <dbl>
## 1  0.23 Ideal     E     SI2      61.5    55   326  3.95  3.98  2.43
## 2  0.21 Premium   E     SI1      59.8    61   326  3.89  3.84  2.31
## 3  0.23 Good      E     VS1      56.9    65   327  4.05  4.07  2.31
## 4  0.29 Premium   I     VS2      62.4    58   334  4.2   4.23  2.63
## 5  0.31 Good      J     SI2      63.3    58   335  4.34  4.35  2.75
## 6  0.24 Very Good J     VVS2     62.8    57   336  3.94  3.96  2.48

#?filter

#filter from dplyr
diamonds %>%
  dplyr::filter(color == "E") %>%
  head()

## # A tibble: 6 × 10
##   carat cut     color clarity depth table price     x     y     z
##   <dbl> <ord>   <ord> <ord>   <dbl> <dbl> <int> <dbl> <dbl> <dbl>
## 1  0.23 Ideal   E     SI2      61.5    55   326  3.95  3.98  2.43
## 2  0.21 Premium E     SI1      59.8    61   326  3.89  3.84  2.31
## 3  0.23 Good    E     VS1      56.9    65   327  4.05  4.07  2.31
## 4  0.22 Fair    E     VS2      65.1    61   337  3.87  3.78  2.49
## 5  0.2  Premium E     SI2      60.2    62   345  3.79  3.75  2.27
## 6  0.32 Premium E     I1       60.9    58   345  4.38  4.42  2.68

Building Graphics

1. Draw by hand (or imagine) the specific plot that you intend to construct
2. Data Wrangling (if needed) to get the data in glyph-ready form, or verify that the current form is glyph-ready for your purposes.
3. Establish the frame using a ggplot() statement
4. Create the intended glyph using geom_[style]() such as
   • geom_point()
   • geom_bar()
   • geom_boxplot()
   • geom_density()
   • geom_vline()
   • geom_segment()
   • geom_histogram()
   • and many more
5. Map variables to the graphical attributes of the glyph using: aes( )

• Rule of thumb: anytime when you are plotting with ggplot, ALL variables need to be inside an aes (except facets, later in slides).

6. Add additional layers to the frame using the + symbol
   • Note: not %>% between layers of ggplot2 graphics
   • Maybe think “add layer” in ggplot2 portions, instead of “and then” with %>% syntax

Steps 4 and 5 can be switched.

https://twitter.com/tanya_shapiro/status/1576935152575340544?t=vwaW8h6CC62h0pkwv9n5Yg&s=19

Example: Baby Names

Let’s look at our BabyNames names data set agian.

# data intake
data("BabyNames", package = "dcData")

# inspect data intake
glimpse(BabyNames)

## Rows: 1,792,091
## Columns: 4
## $ name  <chr> "Mary", "Anna", "Emma", "Elizabeth", "Minnie", "Margaret", "Ida"…
## $ sex   <chr> "F", "F", "F", "F", "F", "F", "F", "F", "F", "F", "F", "F", "F",…
## $ count <int> 7065, 2604, 2003, 1939, 1746, 1578, 1472, 1414, 1320, 1288, 1258…
## $ year  <int> 1880, 1880, 1880, 1880, 1880, 1880, 1880, 1880, 1880, 1880, 1880…

wrangle into glyph-ready form

names <- c("Olivia", "Zoe", "Quentin")

Names <- 
  BabyNames %>%
  filter(name %in% names) %>%
  group_by(name, year) %>%
  summarise(total = sum(count, na.rm = TRUE))

Names %>%
  head()

## # A tibble: 6 × 3
## # Groups:   name [1]
##   name    year total
##   <chr>  <int> <int>
## 1 Olivia  1880    44
## 2 Olivia  1881    51
## 3 Olivia  1882    52
## 4 Olivia  1883    46
## 5 Olivia  1884    54
## 6 Olivia  1885    59

in the beginning you might use mplot to get started–here’s the default result

This isn’t easy to read, and it’s in bad form.

ggplot(data = Names, aes(x = year, y = total)) + geom_line()  + aes(colour = name) + theme(legend.position = "right") + labs(title = "")

we can do better

1. establish the frame
2. plot the glyphs (i.e., select a geom)
3. map the aesthetics
4. add labels and title
5. other features (e.g., alpha, sizing, etc)

Our Plot

1. Establish the Frame

Nothing is here! That is exactly what is supposed to happen. Calling ggplot() only tells us R that we are ready to plot and I want to call some space to create my plot.

ggplot(data = Names) 

2. plot the glyphs (i.e., select a geom)

Still Nothing! We need to tell it what our axis are.

Note that ggplot uses +, NOT %>%. This is because we are adding layers to our plots.

ggplot(data = Names) + 
  geom_line()

## Error in `check_required_aesthetics()`:
## ! geom_line requires the following missing aesthetics: x and y

Note - this is why I like to mape aesthics first, so we can avoid errors.

3. Map the aesthetics

Rule of thumb: anytime when you are plotting with ggplot, ALL variables need to be inside an aes (except facets, later in slides).

#not Quite
ggplot(data = Names) + 
  geom_line( aes(x = year, y = total)) 

#add groups
ggplot(data = Names) + 
  geom_line( aes(x = year, y = total, group = name)) 

#add color
ggplot(data = Names) + 
  geom_line( aes(x = year, y = total, color = name)) 

4. Add labels and title

ggplot(data = Names) + 
  geom_line( aes(x = year, y = total, color = name)) +
  ggtitle("Names Over Time") +
  xlab("Year") +
  ylab("Popularity") +
  guides(color = guide_legend(title = "Siblings Names" ))

5. other features (e.g., alpha, sizing, etc)

ggplot(data = Names) + 
  geom_line( aes(x = year, y = total, color = name, linetype = name)) +
  ggtitle("Names Over Time") +
  xlim(c(1972, 2022))+
  xlab("Year") +
  ylab("Popularity") +
  guides(color = guide_legend(title = "Siblings Names" ), 
         linetype = guide_legend(title = "Still Siblings Names" ))

## Warning: Removed 252 row(s) containing missing values (geom_path).

Remarks about faceting: facet_wrap()

The syntax for facets requires a formula syntax we haven’t seen much yet. Also, there are two main ways to plot with facets. Here are a few pointers:

• facet_wrap() just makes a box for each level of the categorical variable
  • Syntax: facet_wrap( ~ categoricalVariable)
  • For example:

data("NCHS")

# 1is.na(smoker) gets cases that are non-missing for `smoker` (i.e. removes NA's)
Heights <- 
  NCHS %>%
  filter(age > 20, !is.na(smoker)) %>%   
  group_by(sex, smoker, age) %>%
  summarise(height = mean(height, na.rm = TRUE))

head(Heights)

## # A tibble: 6 × 4
## # Groups:   sex, smoker [1]
##   sex    smoker   age height
##   <fct>  <fct>  <dbl>  <dbl>
## 1 female no        21   1.60
## 2 female no        22   1.62
## 3 female no        23   1.61
## 4 female no        24   1.62
## 5 female no        25   1.63
## 6 female no        26   1.62

Heights %>%
  ggplot(aes(x = age, y = height)) +   
  geom_line(aes(linetype = smoker)) +   
  facet_wrap( ~ sex)

Remarks about faceting: facet_grid()

• facet_grid() allows control of row & column facets
• facet_grid() syntax:
  • row & column facets: facet_grid(rows ~ cols)
  • row facets only: facet_grid( rows ~ . ) (note the required “.”)
  • column facets only: facet_grid( ~ cols) (no “.” this time)

Heights %>%
  ggplot(aes(x = age, y = height)) + 
  geom_line(aes(linetype = smoker)) + 
  facet_grid(sex ~ .)

Heights %>%
  ggplot(aes(x = age, y = height)) + 
  geom_line() + 
  facet_grid(sex ~ smoker)

Difference between color and fill

library(mosaicData)

head(CPS85)

##   wage educ race sex hispanic south married exper union age   sector
## 1  9.0   10    W   M       NH    NS Married    27   Not  43    const
## 2  5.5   12    W   M       NH    NS Married    20   Not  38    sales
## 3  3.8   12    W   F       NH    NS  Single     4   Not  22    sales
## 4 10.5   12    W   F       NH    NS Married    29   Not  47 clerical
## 5 15.0   12    W   M       NH    NS Married    40 Union  58    const
## 6  9.0   16    W   F       NH    NS Married    27   Not  49 clerical

CPS85 %>% 
  ggplot() +
  geom_density(aes(x = wage, color = sex), alpha = 0.4)+
  facet_grid( ~ married) +
  xlim(0,30) 

## Warning: Removed 1 rows containing non-finite values (stat_density).

CPS85 %>% 
  ggplot() +
  geom_density(aes(x = wage, fill = sex), alpha = 0.4)+
  facet_grid( ~ married) +
  xlim(0,30) 

## Warning: Removed 1 rows containing non-finite values (stat_density).

CPS85 %>% 
  ggplot() +
  geom_density(aes(x = wage, fill = sex, color = sex), alpha = 0.4)+
  facet_grid( ~ married) +
  xlim(0,30)

## Warning: Removed 1 rows containing non-finite values (stat_density).

CPS85%>%
  ggplot(aes(x = married, color = sex)) + 
  geom_bar() +
  facet_wrap( ~ union, scales = "free")  #Note the scales here 

CPS85%>%
  ggplot(aes(x = married, fill = sex)) + 
  geom_bar()+
  facet_wrap( ~ union, scales = "free")  #Note the scales here 

CPS85%>%
  ggplot(aes(x = age, y = wage, color = sex)) + 
  geom_point()

CPS85%>%
  ggplot(aes(x = age, y = wage, fill = sex)) +  #fill does not work for points!
  geom_point()

Another Example using Diamonds Data

1. establish the frame

2. plot the glyphs (i.e., select a geom)

3. map the aesthetics

4. add labels and title

5. other features (e.g., alpha, sizing, etc)

6. Establish the Frame

ggplot(data = diamonds)

2. plot the glyphs (i.e., select a geom)

ggplot(data = diamonds) +
  geom_point()

## Error in `check_required_aesthetics()`:
## ! geom_point requires the following missing aesthetics: x and y

3. Map the aesthetics

ggplot(data = diamonds, aes(x = carat, y = price)) +
  geom_point()

4. Add Titles and Labels

ggplot(data = diamonds, aes(x = carat, y = price)) +
  geom_point() +
  ggtitle("Diamonds Data") +
  xlab("Carat") +
  ylab("Price")

5. Add additional features

Notice that I can have aes inside multiple statements. Notice that when I use constants (like alpha = 0.3, size = 0.1) they ARE NOT inside aes.

In general, variables go inside aes and constants go outside of it. (unless we are using facets then see previous materials.)

ggplot(data = diamonds, aes(x = carat, y = price)) +
  geom_point(aes(colour = depth), alpha = 0.3, size = 0.1) +
  ggtitle("Diamonds Data") +
  xlab("Carat") +
  ylab("Price") +
  facet_grid( cut ~ color)

ggplot(data = diamonds, aes(x = carat, y = price)) +
  geom_point(colour = "red", alpha = 0.3, size = 0.1) +
  ggtitle("Diamonds Data") +
  xlab("Carat") +
  ylab("Price") +
  facet_grid( cut ~ color)

Side Note about placement of aes

aes can either go inside the ggplot() function, or inside the geom_[chart]() function itself, or both. The 3 following options create the same plots, but the code is slightly different.

#option 1
ggplot(data = diamonds, ) +
  geom_point(aes(x = carat, y = price, color = clarity),
             alpha = 0.2, 
             size = 1) +
  geom_smooth(method = "glm" , 
              formula = y ~ poly(x, 2),                     # y = b_0 + b_1 x + b_2 x^2 + e
              aes(x = carat, y = price), 
              color = "red") +
  ylim(c(0, 20000))

#option 2
ggplot(data = diamonds, aes(x = carat, y = price, color = clarity)) +
  geom_point(alpha = 0.2, 
             size = 1) +
  geom_smooth(method = "glm" , 
              formula = y ~ poly(x, 2),                     # y = b_0 + b_1 x + b_2 x^2 + e
              aes(x = carat, y = price), 
              color = "red") +
  ylim(c(0, 20000))

#Option 3
ggplot(data = diamonds,  aes(x = carat, y = price) )+ 
  geom_point( aes(color = clarity),
              alpha = 0.2, 
              size = 1) +
  geom_smooth(method = "glm" , 
              formula = y ~ poly(x, 2),                     # y = b_0 + b_1 x + b_2 x^2 + e
              color = "red") +
  ylim(c(0, 20000))

• I personally prefer to put “global” aesthetics in the ggplot() and “local” aesthetics in the geom.

  • Option 1 : all aesthetics are local to the geom
    • Note we have to repeat x and y
  • Option 2 : all aesthetics are global to the ggplot
    • Note that color = clarity is not needed for geom_smooth
  • Option 3 : global aesthetics are in the ggplot and local aesthetics and in the geom
    • Both geom_point and geom_smooth use x and y so I put them in the ggplot()
    • Only geom_point uses color = clarity so I put that ONLY in the geom_point function

• In my opinion, Option 3 is the “cleanest” code. This is partly based on stylistic preference and partly based on some internal mechanic of ggplot’s (that is beyond the scope of this course). How you write your code is up to you. Just keep it readable!

• But again, all 3 codes generate the the exact same plot (so does it really matter that much which option we use??)